We are a group of biologists studying the interactions between plants and animals in the rainforests of Panama. Our focus is on the predation and dispersal of seeds by a large rodent called the Agouti. Agoutis prefer large seeds which they can store in scattered underground caches. They later retrieve and eat these caches when there is no fresh food available. This behavior, known as scatter-hoarding, is very important to the trees because it moves the seeds away from the mother tree and plants them into the soil where they can sprout and grow into seedlings. Seeds that remain in the open are generally killed by insects or other seed eating mammals. However, agoutis also have good memories and are able to retrieve and eat most of the seeds they bury.
Agoutis are thus sometimes harmful and sometimes beneficial to trees. This is known as a conditional mutualism and our central question is to determine what conditions affect this relationship. This question is important for understanding forest regeneration, as well as the mutualistic relationships that promote the evolution of diversity.
We are testing two main hypotheses about the control of this mutualism.
- Leftover Food: seeds survive in areas where agoutis have more food than they can use in a year.
- Orphaned Seeds: When an agouti dies, no one knows where the seeds are buried, and these seeds then have a chance to sprout into seedlings.
Seed Tracking: We are using a small motion-sensitive radio-tag to track the fate of seeds. These are attached to large seeds with threads, and allow us to find out where an agouti buries a seed. We then turn off the transmitter to save battery life, and wait for the seed to move again. When the seed is dug up the radio will begin transmitting again, and we will detect this using an Automated Radio Telemetry System (ARTS) that sends live data from towers around our study site to our laboratory on Barro Colorado Island (BCI). We will once again use radio-tracking to find the seed and determine if it has been eaten or moved and re-cached. At the end of the year we will visit all remaining caches to judge if they have germinated.
Agouti Tracking: We are radio-tracking agoutis to determine their home-range areas, and monitor their survival. We use the live data provided by ARTS to detect the death of an agouti quickly, and visit the ‘scene of the crime’ to determine the cause of death. Preliminary analysis suggests an annual mortality rate of 70%, which is typical of a fast reproducing rodent like the agouti.
Food Abundance: We are focusing on the most important food species for agoutis, a large palm known as Astrocaryum. We initially surveyed all of BCI using aerial photographs to obtain a course-scale map of tree density. We then targeted agoutis in the highest and lowest density areas. Once agouti territories are mapped, we map each individual tree from the ground, and count their fruits.
Animal Abundance: We are using camera traps to monitor the local abundance of agoutis, their competitors (e.g. peccaries and squirrels) and their predators (e.g. ocelots and pumas) in each study site. We place 20 cameras randomly within each territory and record photos of passing animals for one week.
Experimental Design: We place tagged seeds in the territories of radio-collared agouti and monitor their removal with a camera trap. Any seed removed and cached by our radio-collared agouti will be monitored long-term with the seed transmitters. If the leftover food hypothesis is correct, seeds should survive and germinate best in areas with high densities of fruit. If the Orphaned Seed hypothesis is correct, seeds should survive and germinate best in the territories of agoutis that die over the course of the study.